bartowski
Qwen2.5 0.5B Instruct
Qwen2.5 0.5B Instruct is a 0.49B dense qwen2 model from bartowski. Configured context length is 32,768 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters0.49B
Context32K
Attentiongqa
Licenseapache-2.0
24 layers · 2 KV heads · head dim 64 · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| F16 | 0.93 GB | Hugging Face file bytes |
| Q8_0 | 0.49 GB | Hugging Face file bytes |
| Q6_K | 0.94 GB | Hugging Face file bytes |
| Q5_K_M | 0.39 GB | Hugging Face file bytes |
| Q5_K_S | 0.38 GB | Hugging Face file bytes |
| Q4_K_M | 0.37 GB | Hugging Face file bytes |
| Q4_K_S | 0.36 GB | Hugging Face file bytes |
| IQ4_XS | 0.33 GB | Hugging Face file bytes |
| Q4_0 | 1.31 GB | Hugging Face file bytes |
| Q3_K_L | 0.34 GB | Hugging Face file bytes |
| Q3_K_M | 0.33 GB | Hugging Face file bytes |
| IQ3_M | 0.32 GB | Hugging Face file bytes |
| Q3_K_S | 0.32 GB | Hugging Face file bytes |
| Q2_K | 0.63 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Runs great | F16 | 145–197 |
| GeForce RTX 3060 12GB | Runs great | F16 | 181–247 |
| GeForce RTX 4060 Ti 16GB | Runs great | F16 | 145–197 |
| GeForce RTX 4090 | Runs great | F16 | 506–691 |
| GeForce RTX 5090 | Runs great | F16 | 900–1228 |
| Apple M4 Pro | Runs great | F16 | 137–187 |
| Apple M5 | Runs great | F16 | 77–105 |
| CPU only · DDR5-5600 | Runs on CPU | F16 | 45–61 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Runs great, F16, 84–115 tok/s
- GeForce RTX 2060 6GB — Runs great, F16, 169–230 tok/s
- GeForce GTX 1660 Super — Runs great, F16, 169–230 tok/s
- GeForce GTX 1660 Ti — Runs great, F16, 145–197 tok/s
- Intel Arc A380 — Runs great, F16, 93–127 tok/s
- GeForce RTX 5060 Ti 8GB — Runs great, F16, 225–307 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (F16).
llama.cpp
llama-server -hf bartowski/Qwen2.5-0.5B-Instruct-GGUF:F16 -c 8192 -ngl 24
Ollama
ollama run qwen2.5
LM Studio
lms get bartowski/Qwen2.5-0.5B-Instruct-GGUF # Then load quant F16